Elimination

Organometallics 1998, 17, 2819-2824

2819

Mechanism of Rearrangement and Alkene Addition/ Elimination Reactions of SiC3H9+ Igor S. Ignatyev† Institute for Silicate Chemistry, Russian Academy of Sciences, Odoevskogo 24, c.2, 199155 St. Petersburg, Russia

Tom Sundius*,† University of Helsinki, Department of Physics, P.O. Box 9, FIN-00014 Helsinki, Finland Received October 15, 1997

Equilibrium geometries of the nine low-energy isomers in the SiC3H9+ system and the transition states for their interconversion have been optimized at the SCF and B3LYP levels of theory. It is shown that the experimentally observed interconversion between the lowest energy isomer (CH3)3Si+ and the second most stable isomer (CH3)(C2H5)HSi+ goes via the intermediate complex RH2Si+‚C2H4. This complex is also an intermediate in the ethene elimination. Due to the key role of the transition state in interconversion between the lowest energy isomer and the complex, in the discussed reactions, its classical barrier height is estimated at different levels of theory. The barrier (∆H298) is found to be in the 4-8 kcal/ mol range above the dissociation limit, in contrast to the case of the SiC2H7+ cation, where it is below the limit. This explains the significant difference in the reaction efficiency for interconversion of the two lowest energy isomers in the SiC2H7+ and SiC3H9+ systems. However, for the SiC3H5D4+ system this barrier for isomerization is predicted to be ca. 4 kcal/mol lower than for the undeuterated species. This may explain the observed complete H/D exchange in the reaction between (CH3)H2Si+ and C2D4. The sequence of transformations of the SiC3H9+ cation which rationalize the mechanism of the observed H/CH3 exchange reaction between SiH3+ and propene has also been found. All interconversions between isomers found in this system are provided only by the hydride shifts. Introduction The experimental study of ethene addition/elimination reactions of silylium cations poses some questions that cannot be readily answered within the existing body of experimental and theoretical data. One of the most intriguing features of these reactions is a dramatic difference in the exchange reactions with labeled ethene between (CH3)2HSi+ and (CH3)H2Si+; the former has only one exchangeable hydrogen, whereas the latter exchanges all five.1-3 The same difference is used to distinguish the SiC3H9+ structural isomers, since (CH3)3Si+ yields only one H/D exchange with C2H4, while (C2H5)(CH3)HSi+ yields seven H/D exchanges.2 Our ab initio analysis of the SiC2H7+ potential energy surface (PES)4 allowed us to suppose that the observed difference in the exchange reactions of deuterated ethene with two types of silylium cations, i.e., RH2Si+ on one hand and R2HSi+ on the other, is connected with the symmetric nature of the RH2Si+‚C2H4 complex. It was proposed that this complex transforms directly into the most stable isomer R(CH3)2Si+ by the simultaneous †

E-mail: I.S.I., [email protected]; T.S., [email protected]. (1) Reuter, K. A.; Jacobson, D. B. Organometallics 1989, 8, 1126. (2) Bakhtiar, R.; Holznagel, C. M.; Jacobson, D. B. Organometallics 1993, 12, 621. (3) Bakhtiar, R.; Holznagel, C. M.; Jacobson, D. B. Organometallics 1993, 12, 880. (4) Ignatyev, I.; Sundius, T. Organometallics 1996, 15, 5674.

shift of two hydrogens from the silylium moiety of the complex to the ethene moiety. Multiple interconversions between the isomers completely scramble H with D. This path is not available for the R2HSi+ cation, and that is why it exchanges only one hydrogen. The above mechanism can take place, provided that the activation barrier for this interconversion can be overcome with the supply of energy provided by complexation with ethene. In other words, the height of the barrier for the interconversion due to the symmetric shift of the two hydrogens in RH2Si+‚C2H4 (TS1) should not exceed the dissociation limit. For the case of R ) H it has been shown that this is true.4 However, it is disputable for R ) Alk. It has been proposed4 that the interconversion between the most stable isomer and the second most stable isomer occurs through the same complex, RH2Si+‚C2H4. Therefore, the barrier height of TS1 relative to the dissociation limit is critical for the successful interconversion of these isomers as well as for the exchange reactions. However, a significant difference between the SiC2H7+ and SiC3H9+ systems was found in the collisional activation study of Bakhtiar, Holznagel, and Jacobson.2 In contrast to the efficient isomerization of (C2H5)H2Si+ into (CH3)2HSi+, only a small amount of (C2H5)(CH3)HSi+ isomerizes into (CH3)3Si+. This observation urged the authors to suppose that isomerization in the SiC3H9+ system invokes a slightly (5-10 kcal/

S0276-7333(97)00898-4 CCC: $15.00 © 1998 American Chemical Society Publication on Web 06/05/1998

2820 Organometallics, Vol. 17, No. 13, 1998

mol) higher barrier for conversion to the most stable isomer than in the SiC2H7+ system and that the barrier in the former case may approach the energy required for ethene elimination.2 Thus, the inefficient isomerization of (C2H5)(CH3)HSi+ into (CH3)3Si+ presupposes the existence of a barrier for their interconversion that is above the dissociation limit, while the complete H/D scrambling in the reaction of C2D4 with (CH3)H2Si+ demands that this barrier should not exceed the energy level of the dissociated molecules. To find an explanation for this contradiction, we have to study the Si(CH3)3+ potential energy surface (PES). Further, there is one more experimental observation that has importance for the proposed mechanism of the (CH3)3Si+ isomer interconversion. It is the H/CH3 exchange reaction between H3Si+ and propene.1 It yields (CH3)H2Si+ and ethene with high efficiency and hence implies small barriers for this reaction. In addition to those problems, which need to be rationalized, the thermochemistry of these cations is of interest in itself, since experimental data are lacking and theoretical studies are limited to only certain isomers of SiC3H9+ at lower levels of theory.5-8 Computational Methods The geometries of the cations were fully optimized at different levels of theory, starting from the self-consistent-field (SCF) level. Two different basis sets were used within this method: 3-21G9 and 6-31G*10 with one set of six d-like functions on Si (R ) 0.45) and C (R ) 0.8). Of the higher theoretical methods which incorporate the electron correlation, the hybrid DFT/HF method was used. This utilizes Becke’s three-parameter exchange functional (B3)11 with the Lee-Yang-Parr nonlocal correlation functional (LYP)12 (B3LYP). This method has been shown to give good results in predicting thermochemical values, including the barrier heights of the reactions (see, for example, refs 1320). The advantage of this method, which becomes more important with the increase of the size of the system studied, is that it is less demanding than the traditional MO methods. The B3LYP method was used with the 6-31G** basis set, in which the polarization p functions are included in the set of the hydrogen basis functions. For the refinement of the relative height of the TS1 activation barrier the diffuse splike (on C and Si) and s-like functions (on H) were used with the 6-31G** as well as with the 6-311G** 21 basis sets. (5) Ibrahim, M. R.; Jorgensen, W. L. J. Am. Chem. Soc. 1989, 111, 819. (6) Cho, S. G. J. Organomet. Chem. 1996, 510, 25. (7) Cremer, D.; Olsson, L.; Ottosson, H. THEOCHEM 1994, 119, 91. (8) Apeloig, Y. In The Chemistry of Organic Silicon Compounds; Patai, S., Rappoport, Z. Eds.; Wiley-Interscience: New York, 1989; Vol. 1, pp 57-225. (9) Gordon, M. S.; Binkley, J. S.; Pople, J. A.; Pietro, W. J.; Hehre, W. J. J. Am. Chem. Soc. 1982, 104, 2797. (10) Francl, M. M.; Pietro, W. J.; Hehre, W. J.; Binkley, J. S.; Gordon, M. S.; DeFrees, D. J.; Pople, J. A. J. Chem. Phys. 1982, 77, 3654. (11) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (12) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785. (13) Durant, J. Chem. Phys. Lett. 1996, 256, 595. (14) Bauschlicher, C. W. Chem. Phys. Lett. 1995, 246, 40. (15) Glukhovtsev, M. N.; Bach, R. D.; Pross, A.; Radom L. Chem. Phys. Lett. 1996, 260, 558. (16) Bell, R. L.; Taveras, D. L.; Truong, T. N.; Simons, J. Int. J. Quantum Chem. 1997, 63, 861. (17) Jursic, B. S. Int. J. Quantum Chem. 1997, 62, 639. (18) Ventura, O. N. Mol. Phys. 1996, 89, 1851. (19) Bernardi, F.; Bottoni, A. J. Phys. Chem. 1997, 101, 1912. (20) Ignatyev, I. S.; Schaefer, H. F. J. Am. Chem. Soc. 1997, 119, 12306. (21) Krishnan, R.; Binkley, J. S.; Seeger, R.; Pople, J. A. J. Chem. Phys. 1980, 72, 650.

Ignatyev and Sundius Due to the importance of the TS1 barrier height for the interpretation of the reaction mechanism in question, two other correlated MO methods were employed solely for studying this point. These include the second-order Møller-Plesset perturbation theory (MP2)22 and the coupled-cluster method including all single and double excitations and all connected triple excitations perturbationally (CCSD(T)).23 In these methods the core-like Si and C orbitals were not included in the correlation. All these methods were used as they are implemented in the Gaussian 94 version of the Gaussian suite of programs.24 The zero-point vibrational energy (ZPVE) was used to obtain the theoretical enthalpy at 0 K from the total energy Ee according to the formula H0 ) Ee + ZPVE. The zero-point vibrational energies were introduced only at the B3LYP level, since this method gives vibrational frequencies in close agreement with experimental ones (see, for example, ref 25), so there is no need for scaling. Theoretical enthalpies at 298 K were calculated, taking account of thermal corrections, from H298 ) Ee + ∆H(298), where

∆H(T) ) Htrans(T) + Hrot(T) + ∆Hvib(T) + RT Both corrections were introduced as they are implemented in the Gaussian 94 codes. The reported ∆H0 and ∆H298 values are theoretical enthalpy changes relative to the most stable isomer. The loosely bound silylium cations have very flat potentials for internal rotations; therefore, they have many rotamers, whose energies, however, usually differ by values not exceeding 0.1 kcal/mol. This problem becomes more important with the increase of the number of methyl groups. While optimizing their geometries, we encountered this problem and tried to locate the lower energy rotamer. However, a complete scan of the rotational potentials was not the goal of this study and we cannot guarantee the absence of lower energy rotamers in some cases.

Results and Discussion Two simple rules evidently govern the relative stability of the silylium cation isomers, and they may be deduced from theoretical studies.4-8,26-31 First, the isomers with a vacant p-orbital (or positive charge) localized at silicon are much more stable than those with a positive charge on carbon. The latter are not actually silylium cations, but rather Si-substituted carbenium cations. Second, the most stable isomer is the one which (22) Binkley, J. S.; Pople, J. A. Int. J. Quantum Chem. 1975, 9, 229. (23) Purvis, G. D.; Bartlett, R. J. J. Chem. Phys. 1982, 76, 1910. Scuseria, G. E.; Janssen, C. L.; Schaefer, H. F. J. Chem. Phys. 1988, 89, 7382. Raghavachari, K.; Trucks, G. W.; Pople, J. A.; Head-Gordon, M. Chem. Phys. Lett. 1989, 157, 479. (24) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.; Johnson, B. G.; Robb, M. A.; Cheeseman, J. R.; Keith, T.; Petersson, G. A.; Montgomery, J. A.; Raghavachari, K.; Al-Laham, M. A.; Zakrzewski, V. G.; Ortiz, J. V.; Foresman, J. B.; Cioslowski, J.; Stefanov, B. B.; Nanayakkara, A.; Challacombe, M.; Peng, C. Y.; Ayala, P. Y.; Chen, W.; Wong, M. W.; Andres, J. L.; Replogle, E. S.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Binkley, J. S.; Defrees, D. J.; Baker, J.; Stewart, J. P.; Head-Gordon, M.; Gonzalez, C.; Pople, J. A. Gaussian 94, Revision B.1; Gaussian, Inc., Pittsburgh, PA, 1995. (25) Wong, M. W. Chem. Phys. Lett. 1996, 256, 391. (26) Hopkinson, A. C.; Lien, M. H. J. Org. Chem. 1981, 46, 998. (27) Wierschke, S. G.; Chandrasekhar, J.; Jorgensen, W. L. J. Am. Chem. Soc. 1985, 107, 1496. (28) Gordon, M. S.; Pederson, L. A.; Bakhtiar, R.; Jacobson, D. B. J. Phys. Chem. 1995, 99, 148. (29) Apeloig, Y.; Karni, M.; Stanger, A.; Schwarz, H.; Drewello, T.; Czekay, G. J. Chem. Soc., Chem. Commun. 1987, 989. (30) Maerker, C.; Kapp, J.; Schleyer, P. v. R. In Organosilicon Chemistry: From Molecules to Materials; Auner, N., Weis, J., Eds.; Wiley-VCH: Weinheim, Germany, 1995; Vol. II, pp 329-359. (31) Pople, J. A.; Apeloig, Y.; Schleyer P. v. R. Chem. Phys. Lett. 1982, 85, 489.

Rearrangement and Reactions of SiC3H9+

Organometallics, Vol. 17, No. 13, 1998 2821

Scheme 1. Geometry and Main Structural Parameters (B3LYP) of the Stationary Points at the SiC3H9+ PES

has the maximum possible number of alkyl groups attached to silicon. This order of stability is similar to that of carbenium cations, i.e., tertiary > secondary > primary. According to the above rules, the least stable isomers should be the primary and secondary silylsubstituted carbenium cations. Even their carbon analogues, such as the 1-butyl cation and the isobutyl cation, were found to represent saddle points rather than local minima at the C4H9+ PES.32 Among the silylium cations, the PES of which was studied at correlated levels of theory, the (SiH3)H2C+ isomer was not found to represent a local minimum at the MP2 level, and this structure rearranges without a barrier to the more stable silylium cation.31 Similarly, there was no energy minimum for the (CH3)H2SiH2C+ isomer at the correlated levels of theory.4 However, the β-silylsubstituted carbenium cation (H3Si)(CH3)HC+ did represent a local minimum, even with the correlated methods.4 This is in keeping with the observation of the destabilizing effect of the R-silyl substitution and the stabilizing effect of the β-silyl substitution relative to the corresponding effects of the alkyl groups.8 Thus, while in our previous study of the SiC2H7+ PES we examined practically all possible isomers of the silylium cation, in the current work we consider only those structures of the carbenium cations which are relevant to the transformations discussed. In keeping with the above rules, the most stable isomer of SiC3H9+ should be (CH3)3Si+ (as in the case of the carbon analogue C4H9+). Indeed, isomer 1 (Scheme 1) was found to represent the global minimum. It has a C3h symmetry with a planar arrangement of the heavy-atom skeleton and with one of the C-H bonds of each methyl group lying in the plane formed by the heavy atoms. This arrangement is similar to that of (32) Siber, S.; Buzek, P.; Schleyer, P. v. R.; Koch, W.; de M. Carneiro, J. W. J. Am. Chem. Soc. 1993, 115, 259.

the lowest energy rotamer of the tert-butyl cation.32 In both cases the rotation of the methyl groups is determined by the compromise between the orientation of C-H bonds, which favors hyperconjugation between the formally vacant p-orbitals on Si (or C) and CH on one hand, and the positions of hydrogens which minimize their repulsion on the other hand. The hyperconjugating C-H bonds are longer than those lying in the plane; however, their difference is not as great as in the case of the tert-butyl cation. The other silicon analogue of the tert-butyl cation, i.e., (CH3)2(SiH3)C+ (7; Scheme 2) is significantly higher in energy (Table 1). The next most stable form should be the secondary silylium isomer, i.e., (C2H5)(CH3)HSi+ (2; Scheme 1). It lies 20.4 kcal/mol above 1 (the ∆H298 B3LYP values relative to 1 will be cited henceforth) and can be regarded as the silicon analogue of the 2-butyl cation. The minimum energy conformation possesses a Cs symmetry. For the primary silylium cations there are two possible isomers which correspond to two forms of the propyl group attached to silicon, i.e., n-propyl (3; Scheme 3) and isopropyl (4; Scheme 2). In the energy gap between 2 and all other forms there are two isomers which can be characterized as alkenesilyl cation complexes: the complex of (CH3)H2Si+ with ethene (5 in Scheme 1, 28.5 kcal/mol) and the complex of H3Si+ and propene (6 in Schemes 2 and 3, 30.4 kcal/ mol). Actually, only 5 may be regarded as a pure complex with a slightly increased ethene CdC bond length of 1.362 Å (B3LYP/6-31G** value in Scheme 1; compare with the ethene equilibrium distance 1.330 Å), with significantly lengthened Si-C bonds (2.339 Å) and negative Mulliken charges on the carbons (-0.26). In 6 silicon is predominantly bound to C2 (Scheme 2), and the charge on C1 becomes positive (0.06). Thus, 6 may be regarded also as a secondary Si-carbenium ion,


Ignatyev and Sundius


Table 1. Total Energy, ∆H (0 and 298 K), for the Most Stable Isomer (Hartrees) and Their Relative Values (kcal/mol) for Other Isomers and Transition States on the SiC3H9+ Potential Energy Surface SCF stationary points 1 2 3 4 5 6 7 8 9 TS1 TS2 TS3 TS4 TS5 TS6 TS7 C2H4 + (CH3)H2Si+ C3H6 + H3Si+ a

B3LYP

3-21G

6-31G*

∆Ee

∆H0a

∆H298

-405.376287

-407.527815

-408.997943

-408.888514

-408.879668

b 0

23.6 45.9 46.8 40.6 45.5 57.0 69.8 67.6 102.7 57.7 73.9 74.6 83.8 82.4 78.2

18.3 36.0 37.7 28.2 30.0 39.6 52.8 48.9 94.4 42.4 53.7 56.4 58.3 61.6 57.4

20.8 37.5 38.7 29.9 33.0 45.9 56.7 50.3 71.2 38.9 53.3 61.9 59.4 63.1 61.7

20.8 37.9 38.8 29.0 31.2 42.4 54.8 49.6 69.2 37.8 50.2 58.1 57.0 59.7 57.9

20.4 36.6 37.7 28.5 30.4 41.9 54.0 48.5 68.3 36.6 49.1 57.1 55.6 58.6 57.0

0 0 0 0 0 0 0 0 1189 690 625 156 374 455 796

72.3 96.6

61.3 80.0

68.4 90.5

64.1 85.9

64.4 86.0

0 0

E0 ) Ee + ZPVE. b Imaginary frequencies. B3LYP values.

although its energy is significantly lower than that of the other Si-carbenium isomer studied here (8) and even the shortest Si-C bond in 6 (2.086 Å) is longer than in the “pure” Si-carbenium ion 8. Two conformations of the heavy-atom skeleton of isomer 8 were found at the SCF level. One of the forms is the cis-like isomer, while the other has a gauche conformation (8; Scheme 3). However, the former does not exist at the B3LYP level; attempts to optimize it at B3LYP/6-31G** lead to hydride transfer from C1 to C2, which results in the lower energy isomer 6. Note that 5 is only 8 kcal/mol above 2 and the activation barrier for its rearrangement into 2 is also only 8 kcal/mol. All isomers of the SiC3H9+ cation may be divided into two classes: one originating from the complex of the

(CH3)H2Si+ cation with ethene (Scheme 1) and the other from that of H3Si+ with propene (Schemes 2 and 3). As in the previously examined case of the SiC2H7+ system, complex 5 transforms into the lowest energy isomer 1 through the transition state TS1 by the symmetric shift of the two hydrogens of the silylium moiety (Scheme 1). This process breaks the C-C bond of the ethene moiety and forms two Si-C bonds, leading to isomer 1, in which all three methyl groups are equivalent. This mechanism explains the exchange of all five hydrogens in the reaction of (CH3)H2Si+ with C2D4. According to this mechanism the complex (CH3)H2Si+‚C2D4 (5) directly converts into (CH3)(CD2H)2Si+ (1) through TS1. The reverse reaction may lead to (CD2H)HDSi+ + C2H3D. Multiple transformations of 5 into 1

Rearrangement and Reactions of SiC3H9+

Organometallics, Vol. 17, No. 13, 1998 2823


and back can completely perform the H/D exchange in (CH3)H2Si+ provided that the activation barrier for this interconversion lies below the dissociation limit. Thus, the height of the TS1 barrier relative to the dissociation limit (∆H298(TS1)) is the critical parameter for the discussion of the two groups of experimental data: (i) the complete H/D scrambling in the (CH3)H2Si+ + C2D4 reaction and (ii) the inefficient isomerization of (C2H5)(CH3)HSi+ into (CH3)3Si+. For (i) the TS1 barrier height should be below the sum of energies of (CH3)H2Si+ + C2D4 and for (ii) the barrier should be above the dissociation limit. ∆E0(TS1) at the B3LYP/ 6-31G** level is 2.8 kcal/mol; ∆H0(TS1) and ∆H298(TS1) are even higher, i.e., 5.1 and 3.9 kcal/mol, respectively (Table 1). If we take into account that the latter value in the SiC2H7+ system is 7.6 kcal/mol below the dissociation limit,4 the overall increase of the barrier height by 12.5 kcal/mol in going from SiC2H7+ to SiC3H9+ is in fair accord with the experimental estimate of 5-10 kcal/ mol2. Thus, according to the B3LYP/6-31G** predictions the classical barrier height for the interconversion between 5 and 1 exceeds the complexation energy by only a few kilocalories per mole and this barrier may be overcome in the reaction of C2D4 with (CH3)H2Si+. Moreover, accounting for the differences in the zeropoint vibrational energy (ZPVE) and thermal corrections between the H3Si+‚C2H4 and H3Si+‚C2D4 systems indicates that the ∆H298(TS1) values for the latter system lie slightly (-0.02 kcal/mol) below the dissociation limit (Table 2). However, it is known from the literature13-20 that the B3LYP method usually underestimates the barrier heights, so it is possible that the closeness of the TS1 barrier height and the dissociation level is an artifact of the B3LYP method. Therefore, we carried out the calculation of the ∆H298(TS1) value for both deuterated and undeuterated systems by other quantum-chemical

Table 2. TS1 Total Energy (Hartree) and the Barrier Height (kcal/mol) Relative to ∆H298 of the Dissociated Fragments C2H4(C2D4) + (CH3)H2Si+ for Undeuterated (∆H298H) and Deuterated (∆H298D) Species method/basis set

Ee

B3LYP/6-31G** B3LYP/6-31++G** B3LYP/6-311++G** MP2(FC)/6-311+G** CCSD(T)/6-311++G**//B3LYP/ 6-311++G**

-408.884 359 -408.886 756 -408.931 613 -408.014 234 -408.090 519

∆H298H ∆H298D 3.8 6.7 7.8 4.0 7.3

0.0 2.9 4.1 0.5 3.6

methods. The studied systems are large enough for the employment of the high level correlated method for their optimization. Such approaches as coupled-cluster theory methods are very accurate and reliable for the prediction of thermochemical properties.33 However, we used the single-point CCSD(T) calculations at the B3LYP/ 6-311++G** optimized geometry. This method gives ∆H298(TS1) values 3.5 kcal/mol above those of B3LYP/ 6-31G** (Table 2). The MP2 values are close to those of B3LYP/6-31G**. Barrier heights in the B3LYP calculations with basis sets which include the diffuse s and p functions are slightly higher, but the scatter of results still does not exceeed 4 kcal/mol with the highest ∆H298D value only 4.1 kcal/mol above the dissociation limit (Table 2). This is within the accuracy of the theoretical methods employed in the prediction of the barrier heights, and we still expect that this barrier can be overcome with the excess complexation energy in the C2D4 + (CH3)H2Si+ reaction. In the above paragraph we discussed the isomerization path from (C2H5)(CH3)HSi+ (2) to (CH3)3Si+ (1), without mentioning the second part of the sequence (33) Lee, T. J.; Scuseria, G. E. In Quantum Mechanical Electronic Structure Calculations with Chemical Accuracy; Langhoff, S. R., Ed.; Kluwer Academic: Dordrecht, The Netherlands, 1995; pp 47-108.


Ignatyev and Sundius

Chart 1. Representation of the Energy Profiles for the SiC3H9+ Isomer Interconversion

depicted in Scheme 1, i.e., the path from 5 to 2. This transformation goes through TS2, which corresponds to the shift of one hydrogen of the silylium moiety of the complex 5 to the ethene moiety of the complex. The barrier for this transformation (TS2) is low (Chart 1) compared to the barrier height of the 5 to 1 transformation (TS1) and thus the entire sequence of transformations from 1 to 2 is determined by the height of the highest barrier (TS1). It is worth noting that in our previous work on SiC2H7+ 4 we did not find a transition state analogous to TS2 (it was found later) and proposed the path leading from 5 to 2 through isomer 4. This path was also found at the SCF level, but it became cut at the B3LYP level, since we failed to find the stationary point corresponding to TS3 within this approximation. The other sequence of transformations (Schemes 2 and 3) has the H3Si+-propene complex 6 as a key structure. Note (as was discussed above) that this structure may be treated as a silyl carbocation as well. In 6 the SiH3 group is in the bridging position toward the CdC bond of the propene molecule, similar to the position of the SiH2(CH3) group toward ethene in complex 5. However, in contrast to 5, where two Si-C distances are equal (2.339 Å at the B3LYP/6-31G** optimized geometry), in complex 6 these two distances are equal to 2.086 and 2.651 Å (6; Scheme 2). The shift of the hydrogen from the central carbon of the propene moiety to the terminal one through TS7 leads to isomer 7, in which the silicon atom becomes coordinated to the central carbon of the propene moiety (SiC ) 1.947 Å). Formally 7 is a Si-substituted tert-butyl cation and thus its relative energy (41.9 kcal/mol) is sufficiently higher than that of the genuine silylium cations. This isomer can comparatively easily (with an activation barrier of ca. 7 kcal/mol) rearrange to the isopropylsilylium cation (4). This rearrangement takes place by the shift of one of the hydrogens of the SiH3 group to the central carbon with the vacant p-orbital (7 f TS3 f 4; Scheme 2).

However, the isopropylsilylium cation (4) has a nonclassical structure with the SiH2 group in the bridging position (Si-C1 ) 1.827 Å and Si-C2 ) 2.222 Å). There is another sequence of transformations of complex 6 (Scheme 3). It starts with hydride transfer (H6) from C2 to C1 through transition state TS4. This transfer results in isomer 8, which is a secondary carbenium ion and therefore it lies ca. 12 kcal/mol higher than the tertiary cation 7. However, the barrier heights for the transformation of 6 to 7 and 6 to 8 are practically of the same value. The carbenium ion 8 also has two possibilities for further rearrangement. The first path leads to the significantly more stable n-propylsilylium cation (3; Scheme 3). More interesting is the second path for the transformation of 8. It goes through the very low barrier of TS5 to an isomer which can be characterized as a complex between ethene and (SiH3)H2C+ (9). This complex may dissociate to the methylsilylium cation (since its carbenium isomer does not correspond to a minimum at the PES31) and ethene. This sequence of transformations of the SiC3H9+ isomers reveals the mechanism of the reaction H3Si+ + C3H6 f CH3H2Si+ + C2H4 observed earlier by Reuter and Jacobson.1 Although it was called formally a H/CH3 exchange reaction, our proposed mechanism shows that the reactions may proceed only by hydride shifts, providing the comparatively low barrier for this process. Note that the predicted exothermicity of the reaction (21.6 kcal/mol) agrees fairly well with that estimated from the thermochemical data by Reuter and Jacobson1 (23.3 kcal/mol). Acknowledgment. I.S.I. thanks the Russian Foundation for Basic Research for supporting this work (Grant No. 96-03-33254). OM970898+

Elimination

Recommend Documents